Advanced analytics can detect and prevent insurance fraud before losses occur

Ydrogios Insurance limits damage, reduces costs and shields its competitive advantage with SAS® Detection and Investigation for Insurance

Since the financial crisis in Greece, fraudsters have become more active, especially in the car insurance sector. This is due to the ease of falsifying or hiding relevant data, creating an enormous financial cost for insurance companies while reducing customer satisfaction and competitive advantage.

A common fraud tactic used in automotive contracts is to try to increase the cost and extent of the damage stated in case of a crash or the shift of liability, to compel the company to cover the claim. Another challenge is identifying connections between injured individuals with the company's customers and whether the injured parties themselves were customers of the company in past years. In both cases, it’s extremely difficult to uncover fraud because the required data is generated from different sources, and the data quality is questionable.

This is why Ydrogios Insurance, one of Greece’s largest insurance companies, has expanded its relationship with SAS. The company found it needed a solution to proactively detect fraud before issuing new contracts to limit damage from falsified claims. As a result, Ydrogios can now offer more cost-effective premiums and more benefits to its policyholders, simultaneously increasing customer satisfaction and reducing financial costs.

SAS Detection and Investigation for Insurance unifies mass amounts of data from internal and external sources. A powerful fraud analysis mechanism then processes all the data in real time. Advanced analytics using artificial intelligence (AI) and machine learning technologies – combined with the use of predictive modeling, text mining, database searches, exception reporting, network link analysis and more – creates a unique end-to-end tool set that recognizes known and unknown patterns in a timely and effective manner.

“It was imperative to invest in a solution that could analyze data from all available channels and sources and provide a unified view of the fraud risk to our company’s investigators,” says Thanos Angelopoulos, Chief Risk and Actuarial Officer of Ydrogios Insurance. “Integrated capabilities for visualization of the results enable understanding and facilitate the decision-making process related to the identification and cost of each damage case.

“We apply the solution in underwriting and in application fraud (falsifying or hiding data during the creation of a new contract application). SAS technology enables us to act proactively by identifying fraud cases on time before the contracts are registered. This is extremely important, because after signing the contract, the damage is hugely expensive and difficult to contain.”

A decrease in insurance fraud and modernization of operations

Using SAS, Ydrogios Insurance increased the detection rate of potential fraud cases from 0.5 percent to 2 percent. At the same time, the company experienced a 5 percent annual decrease in the submission of new claims and a 20 percent reduction in days required to settle a claim and pay compensation to the client.

Angelopoulos also identified some indirect benefits that the company has gained. “Contrary to the general trend of the insurance market, we’ve managed to significantly reduce the frequency of insurance fraud cases and the average cost of claims over the past three years,” he says. “This has enabled us to reduce the levels of premiums of our loyal customers.”

Ydrogios Insurance has significantly increased its competitive advantage. Moreover, the company boosted its credibility and consumer confidence, improved the quality of customer service and reduced the workload for damages settlement. Investigators now have the information visualized in only 10 minutes, compared to past cases when the same results could take two to three days to reach.

“With SAS Detection and Investigation for Insurance, we have the ability to track patterns and potential losses, and to improve the underwriting process by safeguarding the stage of registering new contracts,” Angelopoulos says. “More importantly, the system is constantly improving through machine learning and forces us to become a more data-driven organization, providing the necessary boost to modernize our operations.”